The holidays turned out to be productive, so today there will be four topics. The first is a new community map
abu_hb_lampoviyasd from
abubas. The retrowave atmosphere is almost perfectly recreated in several complex bhop sections, ending with a climb room. The relief on the sides seems to come to life and continuously changes its shape. Magic, no less!
Secondly, duel mechanics have been added to the servers. You can challenge for a duel in the menu
/duel. At that you can call the opponent only if he does not pass the map at the moment, and you yourself must be alive, since the server should check what weapon you have in your hands, as well as what airaccelerate is selected. For example, in this screenshot I see that I can send a request for a duel to one of two players, and the duel will be on famas (I hold it in my hands thanks to the command
/famas), on 100aa (I set it through the command
/aa), with checkpoints allowed (this is set right here in the seventh menu item):
The eighth point allows you to prohibit calling yourself to a duel (this setting is automatically saved in the database along with other settings). Please note that
in a duel you cannot pause, noclip, or go to specs. So before you start, make sure you finish before the map changes. You can leave the duel either by leaving the server, or by doing
/save, or by resetting the timer via
/reset. In this case, the remaining player can complete the map and see a message in the chat about his victory.
Duels are still in test mode, so we are not giving separate rewards for them at the moment. We could, for example, give out Points (not to be confused with Exp), but it’s not entirely clear how to protect duels from farming. Suggest your ideas in the comments, and be sure to PM me if you find any bug.
The third update is
timeout for rtv. We have repeatedly received complaints that a group of players are entering the server, whose goal is to change the map using rtv in order to prevent player currently passing the map from finishing. In this regard, a player who has entered the server will not be able to immediately use rtv if there is at least one player on the server who passes the map and is not AFK (at the same time, players with a running timer now become AFK not after 3 minutes of inactivity, but after 45 seconds). This delay (timeout) is greater the later the player enters the map, but does not exceed a certain limit. For easy maps this limit is 6 minutes, for each next difficulty the limit is one minute more. That is, if I went to an extreme map 20 minutes after it started, and someone passes it, then I will not be able to use rtv for 12 minutes. Of course, if this someone finishes the map or writes rtv himself, then rtv will immediately become available to me.
And the last update for today concerned
skill rating. In the calculations, we already took into account the map difficulty, and the difference between player's record and the world record or the best time on the map (whichever is faster), and we eliminated some of the records so that the rating would not be affected by runs made by the player clearly carelessly (it happens when the map was not good enough, and the player has no desire to improve his time on it). After receiving comments, we not only improved the sifting algorithm by using the calculation of the standard deviation in it, but were also able to take into account the number of player records, so that those who have a couple of dozen good records do not bypass those who have way more good records, but also have some worse records . You can see the result for yourself. The only thing that cannot be taken into account in such a rating is a correct comparison of a player who has many top1 records on easy maps and a player who has many top1 records on difficult maps. In this regard, if you consider it necessary,
lxr can, in addition to the existing one, create two more ratings, with easy and difficult maps, respectively. Write what you think about what we did.
Creep
goeasy
Wooky
We can't share the formula but we will try to make it more accurate.
But I'm not exactly sure about the new Skill Rating.
There are a few ways to compare skill imo
1. Amount of WR demos (XJ and Cosy style)
Pros: You showcase who's dedicated.
Cons: Not a real representation of skill. Also top5 will depend on map pool - if from a thousand maps 900 are blocktime with special techniques (jump bug, tech shortcuts etc.) then expect Cupe / Shooting-star / Xsi to be permanently in the Top 5. On the other hand if those 90% of maps are speed/bhop maps , then they'll probably never even get there (due to their playstyle).
So to make this way of comparing skill work , you have to balance map types in map pool.
And also , when most maps get optimized , newer players will be demotivated to even try , because they'll probably never get a WR until they get somewhat closer to that skill level. (takes years)
Which will bring me to:
2. KZ-Rush points system.
Pros: Allows demo submission of non-WR runs , turning maps into a ranklist.
(for example if you have a 1:28.50 demo on goldbhop , you'll be at Rank 2 for that map , whereas in XJ you'll be ranked nowhere.
Rewards a player for nochecking maps in public server.
Cons: Currently unbalanced because of:
I. 100aa - it should be a separate top, because 100aa is different movement. A separate points system specifically for 100aa, and whichever is higher - appears on your profile. (if you have 12k on 100aa and 5k on 10aa - your VIP level will be for 12k a.k.a VIP4)
II. Difficult maps - it should be reworked imo , because passing with checkpoints hfr_hb_toffifee gives +1 XP , and that takes a lot more skill than nochecking kz_shrubhop_ez (+16 XP). What I suggest is adding a multiplier based on map difficulty. The map I mentioned is the 9th difficulty - Death. So passing it with checkpoints should give 9 XP. While nochecking it should give 9*16 = 144 XP.
3. Playing in real time (cups, duels etc.)
Pros: Shows truly how consistent your skill is.
Cons: Cannot push it to the limit like in demos due to the random nature of the game (fog1/fog2 coinflip).
Can you share how exactly the new Skill Rating is all calculated? Then we will be able to give more decent feedback.
Btw , since 100aa is available on all servers , I suggest turning the only 100aa server into Cosy speed maps server. If you need help with that write me a message, and we'll see what we can do.
Also interesting about "timeout of rtv", I thought it was always a issue of TOO MUCH extending the map, but knowing I didnt play lately, it has to be it changed a lot :D
И найс система дуэли
Hamma colcol 0_0
Regarding the rating system. Overall it is vastly improved from before, well done. I hesitate to give my opinion as it might be seen as criticism when that is far from the case of what I want to do, and obviously it's easier to say something then do it so I really respect the work you guys have put in and the ratings seem a lot better than to what it was previously.
I want to draw the attention to the player jobenz, who is 6th in the rankings (this is not meant as any disrespect to the player whatsoever, his record list in relation to his ranking is simply a good example to elaborate on). Every single top 1 run of his, whether its pro or nub, is a lot slower than the world records of the maps. Furthermore, nearly every single map that he does have a top 1 on, it is clear that from looking at the map rankings there is no competition from any "strong" players, and in the very few cases that they are in the map rankings, it also clear that their time is not up to the standard that they would have if they were to actually spend even a little bit of time on the map. Another point to elaborate on is having nub1 on an easy map that people do pro runs on (which jobenz has) is vastly different from having nub1 on a hard map that very few - if anyone - complete without cp's.
Perhaps a solution to the above would be to rate certain maps more strongly. A suggestion I would give would be to rate a map stronger depending on how many people have actually completed it, as these are usually highly contested maps within the top chain of players, and are popular for a reason. Also, if it would be possible (I don't think it would be viable but just to give an example) to rate maps stronger depending on the level of caliber of players that have actually completed it. Just as an example and from my knowledge as I have a good time on this map, fu_extreme has pretty much the who's who of all hard-extreme map players on the rankings, it should be fair and accurate that one's ranking on this map should be valued very high compared to any other map, as it is clear that if someone can do a great time on a contested map like this, they are also capable of getting a good ranking on any other map of a similar difficulty and similarity as far as what type of map it is (whether its a bhop map, climb or slide map). This could be mis conceptualized by exactly how long a player has spent on the map in relation to the others but I think in general it is a good approach.
Finally, I want to draw the attention of what really makes one consider a player to be better over another (whether in general or in the kz rush rankings)? As an example I'll just name one player of the top of my head, g-Lp. I believe if we were to battle in easy bhop maps he would beat me 10/10 times, yet if we were to battle in easy climb maps with minimal combos, whether its a short or long duration map I am confident I would also beat him 10/10 times. If that was the case, how would we decide who is generally the better player? If we take our abilities on other map difficulties out of the equation, it is a tough question to answer, and so perhaps having separate rankings depending on either the type of maps, and or the difficulties would be most idea.
Sorry for the long text but I have free time and this does interest me :) Everyone should give you guys a big thanks for what you're doing for this game.